Apparatus and method for recognizing road shapes

1. An apparatus for recognizing a road shape of a travel route of a first vehicle, the first vehicle being a vehicle equipped with the apparatus and a surroundings sensor including a camera, the apparatus comprising: a surrounding environment recognizer configured to repeatedly acquire a result of detection by the surroundings sensor and recognize, as surroundings information, at least shapes of roadside objects or travel histories of second vehicles that are vehicles other than the first vehicle, located around the first vehicle in a travel direction of the first vehicle; a sequence-of-point estimator configured to repeatedly estimate, using the surroundings information recognized by the surrounding environment recognizer, a reference line connecting a sequence of points that represent the road shape of the travel route in the travel direction of the first vehicle; and a first lane boundary setter configured to repeatedly set a first lane boundary that is a boundary of a lane of the first vehicle, located at a first distance from the reference line, previously estimated by the sequence-of-point estimator, to a respective one of the left and the right of the first vehicle in a vehicle-widthwise direction of the first vehicle.

2. The apparatus according to claim 1 , further comprising a to-be-excluded vehicle identifier configured to identify, using the travel histories of the second vehicles and first lane boundaries set by first lane boundary setter, a second vehicle, of the second vehicles located around the first vehicle in the travel direction of the first vehicle, whose travel history in the travel direction of the first vehicle does not follow the road shape of the travel route, as a first to-be-excluded vehicle, wherein the reference line is estimated using the surroundings information having the travel history of the first to-be-excluded vehicle removed.

3. The apparatus according to claim 2 , wherein the to-be-excluded vehicle identifier is configured to identify a second vehicle, of the second vehicles located around the first vehicle in the travel direction of the first vehicle, as the first to-be-excluded vehicle when a travel history of the second vehicle in the travel direction of the first vehicle intersects at least one of the first lane boundaries set by first lane boundary setter within a predetermined second distance from the first vehicle in the travel direction of the first vehicle or when the second vehicle is straddling at least one of the first lane boundaries set by first lane boundary setter within the predetermined second distance from the first vehicle in the travel direction of the first vehicle.

4. The apparatus according to claim 3 , wherein a predetermined third distance is equal to or greater than the predetermined second distance, a second vehicle, of the second vehicles located around the first vehicle in the travel direction of the first vehicle, which is located further than the predetermined third distance from the first vehicle in the travel direction of the first vehicle, is referred to as a first spaced-apart vehicle, and the to-be-excluded vehicle identifier is configured to identify a first spaced-apart vehicle as the first to-be-excluded vehicle when a travel history of the first spaced-apart vehicle intersects at least one of the first lane boundaries set by the first lane boundary setter within the predetermined third distance from the first vehicle in the travel direction of the first vehicle, and withhold identifying a first spaced-apart vehicle as the first to-be-excluded vehicle when a travel history of the first spaced-apart vehicle intersects at least one of the first lane boundaries set by the first lane boundary setter further than the predetermined third distance from the first vehicle in the travel direction of the first vehicle.

5. The apparatus according to claim 3 , further comprising a second lane boundary setter configured to set, for each of the first lane boundaries set by first lane boundary setter, a second lane boundary on an opposite side of the first lane boundary from the first vehicle, wherein the second lane boundary setter is configured to acquire the travel histories of the second vehicles recognized by the surrounding environment recognizer, the first lane boundaries set by first lane boundary setter, and information about the first to-be-excluded vehicle identified by the to-be-excluded vehicle identifier, and for each of the first lane boundaries set by first lane boundary setter, when there is a travel history of the second vehicle that is not identified as the first to-be-excluded vehicle on the opposite side of the first lane boundary from the first vehicle, set the second lane boundary located at a predetermined fourth distance from the first lane boundary in the vehicle-widthwise direction of the first vehicle on the opposite side of the first lane boundary from the first vehicle.

6. The apparatus according to claim 3 , further comprising a second lane boundary setter configured to set, for each of the first lane boundaries set by first lane boundary setter, a second lane boundary on an opposite side of the first lane boundary from the first vehicle, wherein the second lane boundary setter is configured to acquire the surroundings information recognized by the surrounding environment recognizer and the first lane boundaries set by first lane boundary setter, and for each of the first lane boundaries set by first lane boundary setter, when there is a lane divider line recognized by the surrounding environment recognizer on an opposite side of the first lane boundary from the first vehicle, set the second lane boundary located at a predetermined fourth distance from the first lane boundary in the vehicle-widthwise direction of the first vehicle on the opposite side of the first lane boundary from the first vehicle where there is the lane divider line recognized by the surrounding environment recognizer.

7. The apparatus according to claim 5 , wherein the second lane boundary setter is configured to, for each of second lane boundaries set by the second lane boundary setter, when there is a travel history of the second vehicle on an opposite side of the second lane boundary from the first lane boundary, set another second lane boundary located at the predetermined fourth distance from the second lane boundary in the vehicle-widthwise direction of the first vehicle on the opposite side of the second lane boundary from the first lane boundary where there is the travel history of the second vehicle.

8. The apparatus according to claim 5 , wherein the second lane boundary setter is configured to, for each of second lane boundaries set by the second lane boundary setter, when there is a lane divider line recognized by the surrounding environment recognizer on an opposite side of the second lane boundary from the first lane boundary, set another second lane boundary located at the predetermined fourth distance from the second lane boundary in the vehicle-widthwise direction of the first vehicle on the opposite side of the second lane boundary from the first vehicle where there is the lane divider line recognized by the surrounding environment recognizer.

9. The apparatus according to claim 5 , wherein the to-be-excluded vehicle identifier is further configured to identify, using the travel histories of the second vehicles and second lane boundaries set by second lane boundary setter, a second vehicle, of the second vehicles located around the first vehicle in the travel direction of the first vehicle, whose travel history does not follow the road shape of the travel route, as a second to-be-excluded vehicle, and reference line is estimated using the surroundings information having the travel history of the second to-be-excluded vehicle removed.

10. The apparatus according to claim 9 , wherein the to-be-excluded vehicle identifier is configured to identify a second vehicle, of the second vehicles located around the first vehicle in the travel direction of the first vehicle, as the second to-be-excluded vehicle when a travel history of the second vehicle in the travel direction of the first vehicle intersects at least one of the second lane boundaries set by second lane boundary setter within the predetermined second distance from the first vehicle in the travel direction of the first vehicle or when the second vehicle is straddling at least one of the second lane boundaries set by second lane boundary setter within the predetermined second distance from the first vehicle in the travel direction of the first vehicle.

11. The apparatus according to claim 10 , wherein a predetermined fifth distance is greater than the predetermined second distance, a second vehicle, of the second vehicles located around the first vehicle in the travel direction of the first vehicle, which is located further than the predetermined fifth distance from the first vehicle in the travel direction of the first vehicle, is referred to as a second spaced-apart vehicle, and the to-be-excluded vehicle identifier is configured to identify a second spaced-apart vehicle as the second to-be-excluded vehicle when a travel history of the second spaced-apart vehicle intersects at least one of the second lane boundaries set by the second lane boundary setter within the predetermined fifth distance from the first vehicle in the travel direction of the first vehicle, and withhold identifying the second spaced-apart vehicle as the second to-be-excluded vehicle when a travel history of the second spaced-apart vehicle intersects at least one of the second lane boundaries set by second lane boundary setter further than the predetermined fifth distance from the first vehicle in the travel direction of the first vehicle.

12. An apparatus for recognizing a road shape of a travel route of a first vehicle, the first vehicle being a vehicle equipped with the apparatus and a surroundings sensor, the apparatus comprising: a surrounding environment recognizer configured to repeatedly acquire a result of detection by the surroundings sensor and recognize, as surroundings information, at least shapes of lane divider lines on a road, shapes of roadside objects, or travel histories of second vehicles that are vehicles other than the first vehicle, located around the first vehicle in a travel direction of the first vehicle; a sequence-of-point estimator configured to repeatedly estimate, using the surroundings information, a sequence of points representing the road shape of the travel route; a recognition distance determiner configured to repeatedly determine a distance from the first vehicle to a most-distant point of the sequence of points estimated by the sequence-of-point estimator as a recognition distance; and a road model calculator configured to repeatedly calculate, using the surroundings information, a road model that represents the road shape with a continuous line, wherein the road model calculator is configured to, when a difference between the recognition distance and a previously determined recognition distance is equal to or less than a predetermined threshold distance, update a previously calculated road model to a currently calculated road model, and when the difference between the recognition distance and the previously determined recognition distance is greater than the predetermined threshold distance, withhold updating the previously calculated road model.

13. An apparatus for recognizing a road shape of a travel route of a first vehicle, the first vehicle being a vehicle equipped with the apparatus, a surroundings sensor, and a vehicle state sensor, the apparatus comprising: a surrounding environment recognizer configured to repeatedly acquire a result of detection by the surroundings sensor and recognize, as surroundings information, at least shapes of lane divider lines on a road, shapes of roadside objects, or travel histories of second vehicles that are vehicles other than the first vehicle, located around the first vehicle in a travel direction of the first vehicle; a sequence-of-point estimator configured to repeatedly estimate, using the surroundings information, a sequence of points representing the road shape of the travel route; a recognition distance determiner configured to repeatedly determine a distance from the first vehicle to a most-distant point of the sequence of points estimated by the sequence-of-point estimator as a recognition distance; and a road model calculator configured to repeatedly calculate, using the surroundings information, a road model that represents the road shape with a continuous line, wherein the road model calculator is configured to, when a difference between the recognition distance and a previously determined recognition distance is equal to or less than a predetermined threshold distance, update a previously calculated road model to a currently calculated road model, and when the difference between the recognition distance and the previously determined recognition distance is greater than the predetermined threshold distance, correct the previously calculated road model using at least a travel speed and a yaw rate of the first vehicle detected by the vehicle state sensor.

14. An apparatus for recognizing a road shape of a travel route of a first vehicle, the first vehicle being a vehicle equipped with the apparatus, a surroundings sensor, and a vehicle state sensor, the apparatus comprising: a surrounding environment recognizer configured to repeatedly acquire a result of detection by the surroundings sensor and recognize, as surroundings information, at least shapes of lane divider lines on a road, shapes of roadside objects, or travel histories of second vehicles that are vehicles other than the first vehicle, located around the first vehicle in a travel direction of the first vehicle; a sequence-of-point estimator configured to repeatedly estimate, using the surroundings information, a sequence of points representing a road shape that is the road shape of the travel route; a recognition distance determiner configured to repeatedly determine a distance from the first vehicle to a most-distant point of the sequence of points estimated by the sequence-of-point estimator as a recognition distance; and a road model calculator configured to repeatedly calculate, using the surroundings information, a road model that represents the road shape with a continuous line, wherein the road model calculator is configured to, when a difference between the recognition distance and a previously determined recognition distance is equal to or less than a predetermined threshold distance, update a previously calculated road model to a currently calculated road model, when the difference between the recognition distance and a previously determined recognition distance is greater than the predetermined threshold distance and a travel speed of the first vehicle detected by the vehicle state sensor is less than a predetermined travel speed, withhold updating the previously calculated road model, and when the difference between the recognition distance and a previously determined recognition distance is greater than the predetermined threshold distance and the travel speed of the first vehicle detected by the vehicle state sensor is equal to or greater than the predetermined travel speed, correct the previously calculated road model using at least the travel speed and a yaw rate of the first vehicle detected by the vehicle state sensor.

15. The apparatus according to claim 12 , wherein the recognition distance determiner is further configured to, when the difference between the recognition distance and the previously determined recognition distance is greater than the predetermined threshold distance, update the recognition distance to the previously determined recognition distance minus a distance that the first vehicle has travelled since when the previously determined recognition distance was determined.

16. A method for recognizing a road shape of a travel route of a first vehicle, the first vehicle being a vehicle equipped with a surroundings sensor including a camera, the method comprising: repeatedly acquiring a result of detection by the surroundings sensor and recognizing, as surroundings information, at least shapes of roadside objects or travel histories of second vehicles that are vehicles other than the first vehicle, located around the first vehicle in a travel direction of the first vehicle; repeatedly estimating, using the surroundings information, a reference line connecting a sequence of points that represent the road shape of the travel route in the travel direction of the first vehicle; and repeatedly setting a first lane boundary that is a boundary of a lane of the first vehicle, located at a first distance from the reference line, that was previously estimated, as to a respective one of the left and the right of the first vehicle in a vehicle-widthwise direction of the first vehicle.

17. A method for recognizing a road shape of a travel route of a first vehicle, the first vehicle being a vehicle equipped with a surroundings sensor, the method comprising: repeatedly acquiring a result of detection by the surroundings sensor and recognizing, as surroundings information, at least shapes of lane divider lines on a road, shapes of roadside objects, or travel histories of second vehicles that are vehicles other than the first vehicle, located around the first vehicle in a travel direction of the first vehicle; repeatedly estimating, using the surroundings information, a sequence of points representing the road shape of the travel route; repeatedly determining a distance from the first vehicle to a most-distant point of the sequence of points estimated, as a recognition distance; and repeatedly calculating, using the surroundings information, a road model that represents the road shape with a continuous line, wherein the repeatedly calculating the road model comprises updating a previously calculated road model to a currently calculated road model for a state in which a difference between the recognition distance and a previously determined recognition distance is equal to or less than a predetermined threshold distance, and withholding updating the previously calculated road model for a state in which gate the difference between the recognition distance and the previously determined recognition distance is greater than the predetermined threshold distance.